The focus of this project is on development and extension of hierarchical modeling approaches for use in genetic epidemiologic cancer research of mutational spectra. In particular,we are developing a hierarchical model that uses information about similarities among mutations (e.g., type, location) to improve estimation of age-specific cancer risk across the spectrum of mutations. We are applying this approach to data on BRCA1 mutations in early onset breast cancer patients and their families, and will be undertaking a simulation study to validate our hierarchical model. This work will provide a coherent approach for obtaining stable mutation-specific risk estimates, and for determining whether risk depends on the type and/or region of mutation (i.e., whether allelic heterogeneity exists). Moreover, the approach developed here will be useful for evaluating other mutational spectra in cancer research (e.g., in the p53 tumor suppressor gene).